Estimation of breeding values

Unbiased data collection guarentees reliable proofs

Germany is a key player regarding the estimation of breeding values. This goes for the traits taken into consideration and the applied estimation methods as well as for the quantity and quality of data. The independent computer center VIT estimates the breeding values for the entire Holstein population according to the latest scientific findings and under supervision of the national board of control. Independent milk recording organizations collect the production data, and only for a few traits, the AI organizations assist in data collection.

Breeding values are estimated for all economically important composites:

• milk production (milk yield, fat, protein)

• functional herd life

• conformation (19 linear traits, 4 scores)

• udder health (somatic cell count)

• fertility (maternal and semen fertility)

• calving ease (daughter calving ease and calving ease)

• calf vitality

• milkability

• temperament

All breeding values are standardized to a relative basis with a mean of 100 and a standard deviation of the true breeding values of 12 points (at 100% reliability). The scale of the relative breeding values is always chosen in a way that the genetically desired stamping of a trait is expressed through a proof of over 100. All relative breeding values refer to a rolling base that is annually adjusted in May and formed by the recent three age groups of young sires.

For the entire Holstein industry, the proofs for most traits are estimated with a BLUP animal model considering all family connections on the basis of the central pedigree data file that contains over 50 million registered animals.

Total Merit Index (RZG)

Already in 1997, the German Holstein Association (DHV) has defined a total merit index (RZG) combining all economically significant composites according to their importance within the breeding goal.

The estimated single trait breeding values are integrated within composites to form relative breeding values. Considering the genetic relations between the composites, they are combined to the RZG with the following weights:

Milk production (RZM)

As the first major Holstein country worldwide, Germany has introduced a test day model already in 1996. Within the scope of a test day model, the evaluation of breeding values is carried out directly on the basis of individual test day records from the official milk recording instead of on the formerly used basis of summarized or projected lactation records. The comparison of the animals within the same herd on the same test day allows the best possible correction for environmental effects. In addition, the test day model reflects individual lactation curves that also provide significant information on all animals’ persistency.

For over 20 years, the records achieved in the first, second, and third lactation are partly treated as independent traits within the evaluation based on a multi trait model. This means that animals with an identical production potential for the first lactation can have a genetically different production level as matured animals (enhancement ability). For a few years now, a multi trait test day model like this one increasingly becomes international standard.

Published is the average breeding value of lactation 1, 2, and 3. In compliance with the breeding goal of a long-lasting, permanently productive cow, the production of the mature cow has a weight of 65% in the published proof.

Since in Germany and Western Europe, the payment for milk is based on the components and especially on the protein content, the proofs for kg protein, kg fat and % protein are summarized in the relative breeding value milk (RZM) with a weighting of 75 : 20 : 5.

Functional herd life (RZN) 

The trait herd life describes the health and constitution of a cow. On account of the significant importance that a long productive life has to the profitability of milk production, this trait has the second highest weight (20%) within the total merit index.

To avoid that a more intensive treatment of high yielding cows (for instance more inseminations, more intensive/longer treatment) results in biased proofs, the trait is defined as functional herd life. The effects of different levels of treatment are corrected by means of production superiority and inferiority respectively within the herd.

The estimation model is based on a lifespan analysis that covers the lifespan of culled animals as well as the present lifespan of animals that are still alive. The proofs resulting directly from this model often have a limited reliability. To enhance reliability, the relative breeding value for herd life not only considers the direct herd life proof but also information from the predictor traits

• udder health (somatic cell count)
• feet & legs (score)
• capacity (body depth)
• udder (udder depth)
• calving ease (maternal stillbirth rate).

The weighting of the predictor traits depends on their reliability and their genetic correlation to herd life. Considering the predictor traits significantly improves the reliability of herd life proofs, in particular for younger bulls. With an increasing number of survival and culling information respectively, the weighting of the predictor traits continuously decreases right until the direct proof for herd life finally is the exclusive source of information for the RZN.

Conformation (RZE)

The data for the estimation of breeding values for conformation are based on the linear assessment and classification of cows and their herd mates in the first lactation in all milk recorded herds. The classified daughters of young sires and the herd mates are randomly and independently preselected by the computer center VIT.

Breeding values are estimated for 19 linear traits as well as for the conformation scores for dairy character, body, feet & legs, and udders. The breeding values for the linear traits are combined to form the composite proofs. These composite proofs and the breeding values for the conformation scores are summarized into the published proofs for the general characteristics dairy character, body, feet & legs, and udder. The relative breeding value conformation (RZE) results from the combination of these four proofs in a ratio of 10 : 20 : 30 : 40.

Somatic cell count (RZS)

In Germany, the somatic cell count is measured at each test day, too. As a result, the modern multi trait test day model can also be applied for this trait.

At higher lactation numbers, an increased somatic cell count as an indicator for clinical or subclinical mastitis can be found more often and with a broader genetic variation between animals. That’s why the proofs for lactation 2 and 3 have a higher weight (37% each) than those for the first lactation (26%). The scale of the relative breeding value somatic cell count is chosen in a way that animals transmitting genetically undesired high cell counts receive breeding values of under 100.

Since the RZS is used as a major predictor trait for the assessment of herd life, the actual influence in the selection is significantly higher than assumed by its direct weight of 7% within the total merit index.

Daughter fertility (RZR)

The estimation of breeding values for the reproductive traits has a long tradition in Germany.

The calculation of fertility proofs is based on the data from all artificial inseminations and all natural services. Quantity and completeness of this database are internationally unique. With a multi trait model, breeding values are estimated for 5 different female fertility traits describing the complexes start of reproductive cycle and conception. The 4 breeding values for the complex conception are combined into the relative breeding value “Conception”. Together with calving to first service (CF) as the trait describing the start of the reproductive cycle, they form the relative breeding value daughter fertility RZR (R = reproduction), with a relative weight of 75% for conception traits and 25% for calving to first service. Daughter fertility has a weight of 10% within the total merit index (RZG).

Calving traits

For calving ease, both the effects of the bull as the sire of the calf (paternal) and as the sire of the dam (maternal calving ability) are estimated. The simultaneous consideration of those 2 breeding values in the selection is important because of the negative genetic link between them.

Calf vitality measured as stillbirth rate and as ‘dead within 48 hours’ respectively is genetically correlated with calving ease, though to some extent it is a separate trait. There are bulls having identical breeding values for calving ease though different values for calf vitality.

In the future, the maternal and the paternal calving traits will be combined in a relative breeding value maternal calving ease and paternal calving ease (RZK mat/pat) respectively. The maternal calving traits have a weight of 3% within the total merit index.

Other breeding values

Besides the proofs included in the total merit index, breeding values for milkability, temperament, and body condition score are estimated, too. The data for milkability (RZD), temperament in the milking parlour (MVH), and body condition score (BCS) result from questioning the owners and recording within the scope of the linear assessment of progeny groups respectively. The multi trait model for milkability also considers milking speed recorded as average milk yield per minute. In the future, BCS proofs calculated within the estimation of breeding values for conformation possibly are due to serve as an auxiliary trait for a reliable estimation of daughter fertility.